1. Field of the Invention
The present invention relates generally to a document-database access device, and more particularly to a method for structuring a URL (Uniform Resource Locators) that is necessary when a WWW (World Wide Web) client accesses a document database, and a document- database access device applying document-database accessing means based on the URL so that the WWW client can follow a hyperlink to retrieve an object document in a case that the document database especially includes a document such as an HTML (Hyper Text Markup Language) document or an XML (Extensible Markup Language) document using a URL as the hyperlink.
2. Description of the Related Art
Japanese Laid-Open Patent Application No. 9-185633 discloses a technology, wherein when a document that includes a hyperlink is uploaded to a WWW server, a group of linked documents are also uploaded to the WWW server. According to the technology, by extracting a URL from the document that is to be uploaded, the document specified by the URL is registered together with the document that includes the URL. The URL can be modified if necessary at the time of an upload process of the documents.
Japanese Laid-Open Patent Application No. 2000-172680 discloses a technology as follows. When a document that includes an URL and a document specified by the URL are uploaded onto a document database, an original URL and a corresponding document identifier in a document database are recorded successively to a correspondence table. In a case that the document is to be taken out from the document database, the document is given to a client after the URL of the document is changed with use of the correspondence table. Though the technology can process any URL, the upload process and a takeout process of the document may take long.
A system using the document database manages contents of the document and bibliography information by relating them with each other, and can retrieve a document under various retrieval conditions. For instance, by setting a document name, an uploading date, a creator name and a word included in the contents of the document as a key, the system using the document database can retrieve the document that fulfills the key conditions.
In the document database, a document may be placed mostly under a folder or a directory. The folders compose a tree structure of the document database, and a necessary document can be retrieved from the document database or can be traced consecutively from the highest-ranked folder to its subfolders. The folders and the documents have names, and when a folder is opened, a list of names of the documents and the subfolders are displayed.
The document database generally includes a server unit and a client unit. The server unit manages document data in the document database. The client unit is operated by a user through a display monitor. A client is a program exclusively used with the document database, and is designed to fit to functions of the document database.
On the other hand, with the propagation of the WWW, it is now necessary to access to the document database not only from the exclusive client but also from a general WWW client, for instance, a WWW browser. In the above-described case, the document database has a function as a WWW server, and a document or a folder in the document database is specified by a URL.
FIG. 1 is a diagram showing one form of a hypertext. The hypertext is logically a single document that is comprised of several linked documents, and is physically divided into several files or nodes. An address of a referred node is specified in order to express a relation from a node to the referred node. For instance in an HTML document, a URL is used for a value of a HREF attribute of an A element or a value of a SRC attribute of an IMG element in order to specify the address of the HTML document.
On the other hand, in the document database, a document cannot be specified only by the file name of the document named before the registration of the document, since different documents might originally have the same file name with the document. Regularly, an identifier ID is given to a single document. The documents and the identifiers are managed in the document database so that the single document corresponding to the identifier is specified if the ID of the document is specified.
FIG. 2 is a diagram showing a situation in which the hypertext shown in FIG. 1 is simply uploaded onto the document database. For instance, a file originally named “hub.html” is managed with a document ID named “doc0001” in the document database. However, the file named “doc0001” includes a URL link address for the other file originally named “note.html”. For instance, assume that the file “doc0001” can be accessed from a WWW client through the following address.                http://hostname/document/doc0001        
Following the above-described URL, the WWW client obtains the contents of the file “doc0001” from the document database and displays the contents on the monitor. Since the file “doc0001” includes a URL link to the file “note.html”, the WWW client interprets the link address as a URL:                http://hostname/document/note.html,and attempts to access the URL. However, the WWW client searches the document with the ID “note.html” through the document database so that the WWW client cannot take out the document corresponding to the file “note.html”.        
To eliminate the disadvantages described above, a document can be specified ID in the document database by using a folder hierarchy and a file name instead of using the document ID. For instance, if the file “doc0001” shown in FIG. 2 can be accessed through a URL:                http://hostname/folder1/folder2/hub.html,the WWW client attempts to access the link through a URL:        http://hostname/folder1/folder2/note.htmlThe WWW client can access the file “note.html” if the files “hub.html” and “note.html” exist under the same folder.        
However, in this document database using URLs, the WWW client must search contents of each folder in order so that a response speed of the document database becomes slow along with a large document-database load. Additionally in a case that documents with a single name exist under a single folder, the WWW client cannot specify the document in the document database using URLs.